FROM THE ENGLISH BOOK
A chamber can pass a leak test yet fail to reach its operating pressure. Line conductance, outgassing and seals in the installed configuration may change the gas balance.
Use pumping speed at the chamber
For a pump connected through a line in the applicable steady-flow model, effective pumping speed is given by:
1/Seff = 1/Spump + 1/C
Seff and Spump are pumping speeds and C is line conductance, all expressed in consistent volume-per-time units. Conductance depends on the gas, geometry and flow regime; use the model appropriate to the operating pressure.
With Spump = 0.10 m³/s and C = 0.010 m³/s, Seff is approximately 0.00909 m³/s. Increasing pump speed to 0.30 m³/s raises effective speed to approximately 0.00968 m³/s, only about 6.5%. The connection is the dominant restriction in this example.
Allocate the gas load before setting a leakage limit
For the adopted steady-state pumping model:
Qleak,lim = Seff(plim − pult) − Qother
Qleak,lim is permissible leakage throughput in Pa·m³/s; plim is the permissible pressure and pult the ultimate pressure used in the model, in Pa. Qother includes the other gas loads in Pa·m³/s. Avoid counting the same gas source both in ultimate pressure and in other gas loads. Use a separate balance for gas components where one effective pumping speed is not justified.
A chamber with limited pumping margin
Let Seff = 0.10 m³/s, plim = 1 × 10−5 Pa and Qother = 8 × 10−7 Pa·m³/s. Neglect ultimate pressure. Total allowable throughput is 1 × 10−6 Pa·m³/s, leaving only 2 × 10−7 Pa·m³/s for leaks. Ignoring other sources overstates the leakage allowance by a factor of five.
A negative calculated leakage budget means the pressure target cannot be achieved under that model even with no leakage contribution. Reduce other gas loads or increase effective pumping speed before assigning a feasible limit.
Keep the conclusion within the test conditions
Pumping speed and leakage throughput have different dimensions. Record gas, pressure, temperature, system configuration and detection checks with the result. An absent detector response establishes only what was not detected under the demonstrated test conditions.
Adapted from English book preview · pages 19–31; original pages 125–137. This online article is an edited summary of selected material.
